# Towards Automatic Concept-based Explanations

arXiv: https://arxiv.org/abs/1902.03129 · HTML: https://mlsanity.com/paper/1902.03129/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 7 Feb 2019.

## Built on by (most cited first, 11 of 11)

- [Bridging the Human-AI Knowledge Gap: Concept Discovery and Transfer in AlphaZero](https://mlsanity.com/paper/2310.16410/) (2023)
- [FP4 All the Way: Fully Quantized Training of LLMs](https://mlsanity.com/paper/2505.19115/) (2025)
- [Robust Semantic Interpretability: Revisiting Concept Activation Vectors](https://mlsanity.com/paper/2104.02768/) (2021)
- [Interpretability of a Deep Learning Model in the Application of Cardiac MRI Segmentation with an ACDC Challenge Dataset](https://mlsanity.com/paper/2103.08590/) (2021)
- [Weakly-Supervised Temporal Action Detection for Fine-Grained Videos with Hierarchical Atomic Actions](https://mlsanity.com/paper/2207.11805/) (2022)
- [Abstraction Alignment: Comparing Model-Learned and Human-Encoded Conceptual Relationships](https://mlsanity.com/paper/2407.12543/) (2024)
- [CliMB: An AI-enabled Partner for Clinical Predictive Modeling](https://mlsanity.com/paper/2410.03736/) (2024)
- [Discovering Concepts in Learned Representations using Statistical Inference and Interactive Visualization](https://mlsanity.com/paper/2202.04753/) (2022)
- [The Standard Interpretable Model: A general theory of interpretable machine learning to deductively design interpretable methods using Lagrangian mechanics](https://mlsanity.com/paper/2606.12289/) (2026)
- [CACTUS: Detecting and Resolving Conflicts in Objective Functions](https://mlsanity.com/paper/2103.07805/) (2021)
- [Model-diff: A Tool for Comparative Study of Language Models in the Input Space](https://mlsanity.com/paper/2412.12177/) (2024)

## Abstract (arXiv)

Interpretability has become an important topic of research as more machine learning (ML) models are deployed and widely used to make important decisions. Most of the current explanation methods provide explanations through feature importance scores, which identify features that are important for each individual input. However, how to systematically summarize and interpret such per sample feature importance scores itself is challenging. In this work, we propose principles and desiderata for concept based explanation, which goes beyond per-sample features to identify higher-level human-understandable concepts that apply across the entire dataset. We develop a new algorithm, ACE, to automatically extract visual concepts. Our systematic experiments demonstrate that alg discovers concepts that are human-meaningful, coherent and important for the neural network's predictions.